Library clips

Buzzlogic seems like a corporate tool to track blog conversations and discover who is influential/authorative.

A post from their blog alludes to some of the mechanisms involved: subscribers, context topic frequency, inlinks, page views and other. They mention how a blog with low inlinks (long tail) can be authorative, but how do they track this:

“Linkless or unattributed influence must be measured, as well. Inlinks are only one way we acknowledge a source of ideas. We’ve found that some sites accelerate individual topics more frequently, presumably because when these authors speak they are treated as more authoritative than others. Because of this a blog with a very low readership deeply interested in a specific topic can be tremendously influential without getting many inlinks. Just consider how frequently we quote others without naming them or attributing the quote to the wrong person, yet the source of the idea has still had influence on the ideas expressed.”

Between The Lines mentions:

“BuzzLogic has a “reach” calculation, which shows how many sites a blog relevantly reaches on a specific subject.”

“In a nutshell, for BuzzLogic an influencer is defined as a “post or publisher generating a significant volume of relevant inbound links and comments about a particular topic or conversation, within a specific timeframe.”".”

Excellent concept on “amplification” from Charbuck.com:

“Mapping a discussion is a very interesting and powerful tool for tracking down the spread of an online meme or rumor. Tracking back to the source of an original concept isn’t very easy. Example: I may detect in a regular blog search, a post that quotes or references a news article or an original blog post. Getting to the source — where the idea was born — isn’t easy, not very precise. But more interesting is the role that “amplifiers” play along the way.”

“Amplifiers may not make the news, but they can give it legs, and in many cases are as important as the original source poster, who may not have the audience and reach, but has the original meme. Case in point, Boing Boing — if Boing Boing points at a blog post, as it did this morning with a blogger who reverse engineered his favorite New York Pizza — they essentially hold up a megaphone to a small voice and instantly turn it into a very loud one.”

So it’s not only calculating popularity, and trying to notice the longtail, but it is really watching for who the influencer is, the person who is the “amplifier”. But aren’t these always going to be the popular blogs (blogs with the highest page view, inlinks, comments, subscribers). These blogs have so many people listening to them that they can amplify anything (almost religious).

DataMining mentions that if, on a given blog, subscribers correlated evenly with inlinks then it seems fair to say we could rely on inlinks as a measure of influence. That is “influence opportunity” (someone may read your post and may decide to link to it in their post) can only actualize if you have subscribers (someone to read your post)…or discover you in search results (which you won’t rank high without being popular).

But Matt’s analysis on 800 blogs shows that it doesn’t correlate that strongly, from the graph we can see some blogs with 100 subscribers have 1000 inlinks, and other blogs with 10000 subscribers only have 50 inlinks.

So it is not always the case the blogs with high inlinks will be more influential than blogs with low inlinks, only of course if these blogs with low inlinks have a high subscriber count. If we say that a blog with a high subscriber count is popular, how do we determine this for a topic (perhaps then we look at topic/context analysis).

I’d like to see this analysis take into account the frequency of blog comments and then merging them into one graph. Measuring inlinks, subscribers, comments, page views, bookmark frequency and tags used (how many times any of a blogs posts have been bookmarked by the various bookmark services), tags people use when they bookmark posts may give a clue on topics a blog is seen to be an influencer (also consider the tags or category a blogger users to organise their posts as a measure of what they think they are experts in)…then from these criteria, along with context analysis and word frequency we can discover bloggers who “amplify certain posts” and over time see if this develops into the blogger “amplifing” posts on a certain topic, thus being an influence on that topic, perhaps driving the blogosphere on that topic.

Expert vs Amplifier

We have to be careful to see the difference between an “expert” and an “amplifier”. An amplifier is popular (thus is an influencer) and spreads the word, everyone is waiting for them to say something from which the message spreads.
An amplifier doesn’t even have to break the news, they can always be getting their news from another source, which for some reason isn’t as popular, perhaps this source has the latest news but doesn’t write compelling posts, so they don’t attract subscribers, or perhaps it is hard for them to break into a saturated market even though they have the gossip first, as people are happy and trust another proven blog that also is attractive because it has a community feel (lots of comments).
So even though blog A may break the news before blog B everytime, this may make them the source but not the amplifier, and really whose source are they, they are only the source to a few that know about them, to everyone else the amplifier is the one and only source.

An expert (who can also be an amplifier), isn’t necessarily an amplifier. Any expert or an authorative person on a topic may post infrequently, but when they do, it’s a gem…but I guess they have to post enough of these gems to even prove (maybe remind) to themselves they are an expert on a topic.
Whether people subscribe or link to you may be irrelevant, you know you are an expert on this topic outside of the blogosphere, which means when you publish your know-how on in the blogosphere it is coming from an expert, but the blogosphere doesn’t recogise you are one, you have to be visible.

eg. you may be the best TV repair man on the planet, but if you don’t make youself visible no-one will know about you and you don’t make money, in the blogosphere this metaphor refers to popularity.
(although, there are situations where someone who is quiet is still so enigmatic that they draw people to them without trying, and once an “amplifier” knows about you then you can capitalise on your expertise/authoritativeness).

A lone shaman blogger could be the most authoritative person on plant knowledge in a region, but instead a scholar blogger is known as the most authoritative, because he belongs to a sub-culture, an incrowd, he has blogger friends, and once a couple of A-listers form a posse on a given topic, it’s hard to get in as an amplifier or expert.

So it is not just about posting, but it’s about getting involved, contributing with comments, subscribing and linking to people, IM and email them, whatever, get part of a community and soon enough this is recipricated and you maybe be known as a proven expert or perhaps become an ampifier (if this is what you want, it will be the by-product of you enthusiasm anyway).

More on subscribers

Maybe some blogs with high readership are so formal that they don’t really write provoking posts, or induce a conversational environment, but people follow them because they are professional and perhaps write formal review type posts (and as I said before have proven themselves over a long time, they may have been a pioneer). If the content doesn’t invoke people to add to it, then you won’t be referenced much (although, you may be referenced a lot if you are breaking news). But this doesn’t stop you from being an expert, as you have a loyal readership.

Some blogs with low readership may have lots of links in comparison because their blog is old in comparison (it has had a chance to accumulate these links over younger blogs), so I think age of the blog has to be included, perhaps inlinks and subscriber count measured over the last year.
Maybe these blogs have high inlinks compared to subscribers because every once in a while they post some breaking news that an A-List blogger will “amplify” and in turn generate some link love (if the amplifier has chosen to reference them). The readers of this A-List blog maybe not subscribe to the source blog as it might not post often, or its content is not usually the same as the content of the post that was linked to a lot (in this case they might want to look at their stats and write posts that people seem to like).

So it may be the fact that because you are on an A-listers Reading List, your occasional post gets amplified generating you some links. Posting once a week, may get you more inlinks than someone posting 10 times a day, even though the quality of your posts and subscriber numbers are the same.

It may be the quality of your subscribers that matter, if your total suscbribers is just 5 A-listers and your mom, and an equivalent blog has 1000 subscribers, you may still get more inlinks

…sometimes it’s who you know.

Some notable resources

BlogPulse Conversation Tracker
Technorati
Sphere
Talkdigger
TechMeme
megite
TailRank
Wizag (formerly Diggol)
Personal Bee
MyFeedz
Feeds 2.0
The Truth Laid Bear

My posts on this subject

Talkdigger conversations and more
Sphere and others : related posts from the blogosphere
Bloglines blog search engine
Distributed conversations: pinging and tagging
Blog search: context relevancy
Blog Ranking: Incoming links??
Blog Clouds
Diggol : personalised RSS Reader with the lot
Personal Bee: Tag cloud RSS reader
MyFeedz : attentive news by tag
Feeds 2.0 : personal memetracking reader

So I heard Google Reader has a trends feature, the basic premise is that it is recording the feeds I pay attention to and letting me know about it, what it isn’t doing is actioning on that, it will let me decide that.

What I mean is that it is not going to delete feeds I don’t read often or feeds that don’t post frequently or reshuffle the rank/order of my feeds, it’s just letting me know.

I am really excited about this feature, as it is telling me my top 10 (based on stories I read), this is kind of bias if a feed you read doesn’t post that often…but it is giving more than I need to make some feed management decisions.
We all have feed overload and management issues, so this is a handy tool to crop stuff out, and I won’t feel sad to let go, as I know based on activity that it’s time to say bye. In contrast, it may be time to reshuffle feeds, create new folders, etc…

NOTE: You can set-up your feed reading to auto-sort, ie. set it so less active feed items are on top.
You can also sort your items to read them in reverse, ie. starting with the oldest items

Features

Summary - eg. From your 303 subscriptions, over the last 30 days you read 859 items, starred 6 items, and shared 0 items.

Reading Trends - Top 10, 20, and 40 feeds based on the number of items “read” (also gives %)
It also gives these statistics for items “starred” and items “shared”, basically this is an aspect of personalisation.

Tag cloud - how many items a tag has the bigger it is, the more I’ve read the darker it is
Tags are folders you organise your feeds in (the output of these can be automated into a public stream)…you can also manually tag items

Chart - a bar chart of the number of items read in the last 4 weeks (a bar for each day)…also shows which day of the week I read a lot or which days I don’t read at all eg. the weekend is always zero for me…also shows what time of day I read the most and vice versa.

This chart not only tells me about my feed reading, but tells me about my productivity/time management (I may look at this chart and say to myself do I do anything else with my time but read feeds…

Subscription Trends - Top 10, 20, and 40 frequently updated feeds based on items per day, and shows the Inactive feeds (giving the last updated date).
Basically I can get a picture of the heavily posting feeds, which I have done and put them in their own folder, I can also see the feeds that are inactive and perhaps delete them or put them in a dead folder.

What’s next!

Recommendation - I can see this information from all users being aggregated to recommend you feeds and items, eg. “here are people that clip the same items as you, or read a lot of the feeds you do, so go check out their user page, and easy add feeds, clip blogs, add the feed of their whole account, clip items, etc…”

In turn there will be a feed folksonomy (or even something like SYO) and an item folksonomy, as well as both an item and feed Community Top 10 (also by topic)…basically the inevitable progession like Rojo to a top buzzing stories stream with perhaps memetracking to boot, or even memedigging by aggregating the “shared” feature.

Social network - connect the system so you can find people and view a users RSS Reader, view clip blogs and see all this stuff in aggregrate…also easy add as a friend button, and easy add clip blogs button.

Wishlist

- manually order the feeds in a folder, and I wish I could manually order the folders (also you can’t rename folders)

- collapse similar stories like Google News and Feeds2.0, this way I can read the lastest buzz around a post in one clean sweep, instead of in chunks as I go through my reader (basically a type of personal memetracking).

This can also be done by viewing content by wordburst tags.

- search (limit by folder, limit by tag…even be able to select feeds or folders to search)…more

- see comments for a post and be able to post a comment from within Google Reader, then perhaps track the conversation ensuing within Google Reader instead of using a 3rd party like co.mments.

- a scratch pad like Gritwire, or maybe being able to send a post to Google Notebook, now that would be awesome as one of my notebooks is used to collect links for future posts, each section heading in this notebook is a future post. I wish I had a bookmarklet for this to add stuff (to a section heading within a notebook) while I’m surfing the web, the extension does the trick I suppose. As I said, sending an item to a section heading in a notebook from within Google Reader would be the bomb.

- another view of your public streams from your non-folder tags (just your manual tags)…another view would be like a traditional bookmark view

- unique export OPML, ie. an OPML with my user name in the URL, also the same for each folder

- include attention data in the OPML, perhaps a dynamic (always changing) OPML for your Top 10 Reading Trends…kind of like a http://libraryclips.blogsome.com/2006/08/28/grazing-conversations/“>feed grazer (subscribe or import your current Top 10 Reading Trends OPML, or subscribe to its spliced feed).

Imagine a Top 10 grazing list for the whole of the Google Reader community, great for feed shopping, or even a grazing list of feeds from the Top 10 items…more and more.

- clip blogs to look like a real blog

- desktop version that synchs (now that’s asking a bit much)

Check out the wishlist at Web Worker Daily.

I’m really convinced that Google Reader is here to stay, right now there’s no better…a lot of readers have their time in the spotlight, but Google Reader is doing everything right, especially the fundamental thing of reading feeds with ease in the shortest time possible…not sure if anything will outshine it.

Maybe my wishlist will be more well received in the forum.

Related:
Google Reader, clipmarks and SuprGlu splice up
Google Reader and clipping in general