Blogroll | Communications of the ACM

From The Noisy Channel

Evaluating Search: Using Human Judgement

Originally posted on the Twiggle blog.In the previous post, we looked at measuring searcher behavior in order to evaluate search engine performance. Measuring searcher... Daniel Tunkelang
From The Noisy Channel | October 1, 2017 at 10:08 PM

From The Noisy Channel

Evaluating Search: Measuring Searcher Behavior

Originally posted on the Twiggle blog.Measuring the effectiveness of your search engine is hard. Fortunately, you have an army of volunteers helping you: your customers... Daniel Tunkelang
From The Noisy Channel | October 1, 2017 at 10:01 PM

From The Noisy Channel

Evaluating Good Search (Part I): Measure It

Originally posted on the Twiggle blog.At Twiggle, we’re all about improving the search experience. But how do we define improvement? As Lord Kelvin, one of history... Daniel Tunkelang
From The Noisy Channel | October 1, 2017 at 09:15 PM

From The Noisy Channel

Contextual Query Understanding

So far, we’ve focused on understanding searchers based entirely on the words they type into the search box. But search doesn’t occur in a vacuum. In the next posts... Daniel Tunkelang
From The Noisy Channel | September 11, 2017 at 08:12 PM

From The Noisy Channel

10 Things Everyone Should Know About Machine Learning

As someone who often finds himself explaining machine learning to non-experts, I offer the following list as a public service announcement. Machine learning means... Daniel Tunkelang
From The Noisy Channel | September 5, 2017 at 01:34 PM

From The Noisy Channel

Autocomplete and User Experience

The previous post focused on how to determine the best autocomplete suggestions based on query probability and query performance. In this post, we’ll dive intoscope... Daniel Tunkelang
From The Noisy Channel | September 1, 2017 at 05:03 PM

From The Noisy Channel

You’re absolutely right about the number of suggestions, of course.

You’re absolutely right about the number of suggestions, of course. And it’s a good reminder that autocomplete user experience deserves its own post, which I promise... Daniel Tunkelang
From The Noisy Channel | August 28, 2017 at 05:07 PM

From The Noisy Channel

Autocomplete

In the past decade, autocomplete has become a required feature for search engines. Today, searchers who type into a search box expect to see autocomplete suggestions... Daniel Tunkelang
From The Noisy Channel | August 28, 2017 at 10:18 AM

From The Noisy Channel

I agree with the judge about the balance of hardships, but that in itself wouldn’t have been…

I agree with the judge about the balance of hardships, but that in itself wouldn’t have been sufficient to grant the injunction. The judge also ruled that “hiQ... Daniel Tunkelang
From The Noisy Channel | August 18, 2017 at 01:15 PM

From The Noisy Channel

On hiQ v. LinkedIn

On August 14th, US District Judge Edward M. Chen granted a preliminary injunction to hiQ Labs in its case against LinkedIn. While I agree with part of the judge... Daniel Tunkelang
From The Noisy Channel | August 16, 2017 at 01:08 AM

From The Noisy Channel

Here are a couple of relevant papers:

Here are a couple of relevant papers: Scaling Semantic Parsers with On-the-Fly Ontology Matching Ontology-Based Translation of Natural Language Queries to SPARQL... Daniel Tunkelang
From The Noisy Channel | July 27, 2017 at 04:36 PM

From The Noisy Channel

Taxonomies and Ontologies

In order to understand queries, it’s important to ground that understanding in a knowledge base. Two common ways to represent a knowledge base are taxonomies and... Daniel Tunkelang
From The Noisy Channel | July 27, 2017 at 03:51 PM

From The Noisy Channel

Prithiviraj, thanks for the kind words.

Prithiviraj, thanks for the kind words. As for NER systems for English text only recognizing named entities in title case, that’s usually a function of how they... Daniel Tunkelang
From The Noisy Channel | May 31, 2017 at 11:15 AM

From The Noisy Channel

Entity Recognition

In the previous post on query scoping, we discussed query tagging as a special case of named-entity recognition (NER). In this post, we’ll dive a little bit deeper... Daniel Tunkelang
From The Noisy Channel | May 30, 2017 at 09:31 AM

From The Noisy Channel

Query Scoping

In the previous post, we discussed how query segmentation improves precision by grouping query words into semantic units. In this post, we’ll discuss query scoping... Daniel Tunkelang
From The Noisy Channel | May 16, 2017 at 11:46 AM

From The Noisy Channel

Thanks Bart. I haven’t read it, but I’ll take a look. Thanks for sharing.

Thanks Bart. I haven’t read it, but I’ll take a look. Thanks for sharing. Daniel Tunkelang
From The Noisy Channel | April 29, 2017 at 04:27 PM

From The Noisy Channel

Query Segmentation

The previous two posts focused on using query rewriting to increase recall. We can also use query rewriting to increase precision — that is, to reduce the number... Daniel Tunkelang
From The Noisy Channel | April 28, 2017 at 02:51 PM

From The Noisy Channel

A Non-Adversarial Ad-Supported Model

If you’ve followed me for a while, you know that I’m not a big fan of ad-supported business models. It’s not just that I find ads annoying; I also see them as economically... Daniel Tunkelang
From The Noisy Channel | April 15, 2017 at 07:23 PM

From The Noisy Channel

Thanks for the kind words!

Thanks for the kind words!Stack Overflow and Quora certainly have reputation systems, and I love both sites. But I don’t feel either of them is focusing on being... Daniel Tunkelang
From The Noisy Channel | April 6, 2017 at 02:32 PM

From The Noisy Channel

How could LinkedIn be disrupted?

When I worked at LinkedIn, I focused on its mission of “connecting talent to opportunity at a massive scale”. Or, as I preferred to describe it, helping people... Daniel Tunkelang
From The Noisy Channel | April 5, 2017 at 06:59 PM